Research paper
Midbrain responses to micro-stimulation of the cochlea using high density thin-film arrays

https://doi.org/10.1016/j.heares.2012.04.004Get rights and content

Abstract

A broader activation of auditory nerve fibres than normal using a cochlear implant contributes to poor frequency discrimination. As cochlear implants also deliver a restricted dynamic range, this hinders the ability to segregate sound sources. Better frequency coding and control over amplitude may be achieved by limiting current spread during electrical stimulation of the cochlea and positioning electrodes closer to the modiolus. Thin-film high density microelectrode arrays and conventional platinum ring electrode arrays were used to stimulate the cochlea of urethane-anaesthetized rats and responses compared. Neurophysiological recordings were taken at 197 multi-unit clusters in the central nucleus of the inferior colliculus (CIC), a site that receives direct monaural innervation from the cochlear nucleus. CIC responses to both the platinum ring and high density electrodes were recorded and differences in activity to changes in stimulation intensity, thresholds and frequency coding of neural activation were examined. The high density electrode array elicited less CIC activity at nonspecific frequency regions than the platinum ring electrode array. The high density electrode array produced significantly lower thresholds and larger dynamic ranges than the platinum ring electrode array when positioned close to the modiolus. These results suggest that a higher density of stimulation sites on electrodes that effectively ‘aim’ current, combined with placement closer to the modiolus would permit finer control over charge delivery. This may equate to improved frequency specific perception and control over amplitude when using future cochlear implant devices.

Highlights

► Responses to cochlea stimulation using two electrode designs were compared. ► Thin-film electrodes produced more localised frequency responses. ► Thin-film electrodes produced lower threshold and larger dynamic range. ► Results indicate thin-film designs may provide enhanced sound perception.

Introduction

Cochlear implants (CI) can provide open-set speech perception to prelingually and postlingually deafened individuals and the ability to converse via the telephone (Brown et al., 1985; Clark et al., 1981; Dawson et al., 1992). However, elements of the auditory scene required for speech perception in background noise or for full appreciation of music are not available to CI users (Clark, 2003; Driscoll et al., 2009; Moore, 2007). The perception of speech in background noise requires sufficient auditory cues and the listener's ability to discriminate between, and predict small changes in frequency and amplitude (Bavin et al., 2010; Jerger and Musiek, 2000; Lindblom et al., 2009; Näätänen and Winkler, 1999; Osberger et al., 1991; Tobey, 1993; Zatorre et al., 1999). The perception of music requires discrimination between and detection of a wide range of frequencies, timbres and amplitudes for sufficient comprehension to occur (Deutsch, 1991; Patel, 1998; Strait et al., 2010). Presently CI devices can provide only a restricted number of discrete frequencies and a limited dynamic range (Cohen et al., 2001; Drennan and Rubinstein, 2008; Shepherd et al., 2004), hindering the ability to comprehend complex sounds.

There are several categories of hearing loss and individuals with profound sensorineural hearing loss require intensities of 90 dB or above to perceive sound (Boothroyd, 1993; Shalit and Avraham, 2008). In all cases of deafness, auditory perceptual thresholds are increased and the ability to comprehend speech is compromised (Moore, 2007). For those with normal hearing the perception of acoustic environments requires the discrimination of several sources of sound that can have discreet fluctuations in frequency, intensity, timing and place (Lindblom et al., 2009; Lutfi, 2008; Moore, 2007; Plomp, 1976). Psychophysical aspects of frequency perception observed in the normal hearing population are fundamental to complex sound perception and as such the delivery of discrete frequency input and a greater control over amplitude should be goals in the development of CIs.

Although sound waves activate discrete regions of spiral ganglion neurons in the cochlea, typical CI use activates broad areas of auditory nerve fibres beyond specific frequency regions, particularly at high frequency areas (Drennan and Rubinstein, 2008; Lim et al., 1989; Middlebrooks and Snyder, 2007; Moore, 2007). The presumed size of the electrical field during stimulation of the cochlea has been shown to correlate with auditory cortex activity, indicating that limiting electrical spread may improve frequency perception (Bierer and Middlebrooks, 2002; Bonham and Litvak, 2008; Clark, 2006). Perilymph and bone in the cochlea impede the ability to limit current to specific frequency regions and research has shown that direct stimulation of auditory nerve fibres can elicit more localised frequency responses and thresholds at lower currents than traditional CIs (Middlebrooks and Snyder, 2007). Furthermore, the number of distinct frequencies that users can perceive is linked to the number of active channels in the cochlea (Xu and Pfingst, 2008). Therefore, increasing the number of available and perceptually distinct electrode sites in combination with a decreased field of electrical activity during stimulation may improve the efficacy of CI listening (Clark, 2006: Wise et al., 2008).

The ability to perceive both speech and music are not believed to be separate cognitive modalities (Deutsch, 1991; Patel, 1998). Speech processing strategies can be used to enhance either temporal cues at the expense of spectral cues or vice versa (Drennen et al., 2010) and it has been argued that the development of a new generation of CIs does not necessarily require the inclusion of more distinct active electrode sites (Jung et al., 2009). However, CI users' perception of higher frequency changes (areas of the cochlea where frequency changes cover smaller regions of spiral ganglion neurons; Moore, 2007) is poor (Lim et al., 1989), and more discrete activation in these areas may allow for better spectral resolution. If spectral input can be enhanced through electrode design then temporal properties may not need to be the trade-off.

Detecting changes in amplitude are important in determining the location of sound sources (Hu and Wang, 2004). Electrode arrays placed closer to the modiolus can lower the thresholds required to elicit neural activity in higher order processing sites of the auditory pathway during CI stimulation (Shepherd et al., 1993). Distance of electrode arrays from the modiolus also negatively correlates with psychophysical and physiological dynamic range, and accordingly it has been argued that future designs should attempt to minimise this distance (Cohen et al., 2001; Shepherd et al., 2004). Through increased dynamic range and lowered thresholds close modiolus proximity has the potential to improve sound perception using CIs.

The aim of this research was to use an experimental thin-film high density electrode array (HDA) comprised of 32 sites and compare neurophysiological response with that elicited by a conventional platinum ring electrode array covering a similar length of the cochlea. Small iridium electrode sites on the HDA can provide a greater charge density than the traditionally used platinum rings (Beebe and Rose, 1988; Robblee et al., 1986) and the potential to decrease threshold and increase dynamic range. Recordings in the central nucleus of the inferior colliculus (CIC), which receives direct monaural innervation via the contralateral cochlear nucleus (Winer and Schreiner, 2005), of hooded Wistar rats, were used to compare frequency specific activity, threshold and dynamic range of the electrode array types. A greater density of electrode sites offers the ability to limit linear distance between stimulation and reference sites and as such, the HDA has the capability to minimise the field of electrical stimulation, offering better frequency specific activation. We anticipated that proximity of electrode arrays to the modiolus would be correlated with lower thresholds and larger dynamic ranges in the CIC and that the HDA would produce lower thresholds and larger dynamic ranges than the platinum ring array. Findings of this research may offer a path to enhanced sound perception using CIs.

Section snippets

Surgery

Male adult Hooded Wistar rats (n = 9), weighing between 200 and 500 g, were used as subjects. Animals were anaesthetised prior to surgery via intraperitoneal administration of urethane/distilled-water solution (20/80, 1.5 g/kg). During surgery, normal body temperature (36.5 °C) was maintained using a DC-homoeothermic blanket. Animals were randomly assigned to either a control group (n = 4) or an experimental group (n = 5), which determined the CI electrode used. All procedures and protocols

Recording electrode placement and characteristic frequencies

Data was collected from 197 multi-unit recording sites in the CIC. An outline of the number of recording electrodes and frequency range for each animal is shown in Table 1. Histology was used to confirm recording locations. Recording electrodes showed response patterns and tonotopicity that confirmed CIC placement. To confirm that driven neural response was a result of acoustic or electrical stimulation, PSTHs were generated. Recording electrode placements for all animals and examples of (both

Discussion

The aim of this research was to establish whether HDAs could elicit more localised frequency-specific neural activation, lower electrically evoked thresholds and larger dynamic ranges in the CIC compared with a more traditional platinum ring array. Regardless of the spacing between stimulation and reference sites, the thin-film electrode elicited more specific CIC frequency responses than the platinum ring array. The HDA produced a lower CIC threshold and significantly larger dynamic range

Conclusions

HDAs appear to have the potential to provide an enhanced sound experience in comparison to conventional CI designs. However, the placement of HDAs has an impact on the functionality. This is consistent with previous research that found that although arrays with smaller electrode sites could elicit lower thresholds and larger dynamic ranges, these benefits were present only in combination with precise electrode placement (Clark et al., 1983). Results suggest that the design of the HDA has

Acknowledgements

This research was funded by the ARC Centre of Excellence for Electromaterials Science. The authors wish to thank Rosalia Bruzzese for her technical support. We also wish to thank Peter Allitt and Matt Pennell for proof reading.

References (55)

  • K. Wise et al.

    High-density cochlear implants with position sensing and control

    Hearing Research

    (2008)
  • L. Xu et al.

    Spectral and temporal cues for speech recognition: implications for auditory prostheses

    Hearing Research

    (2008)
  • R.J. Zatorre et al.

    Auditory attention to space and frequency activates similar cerebral systems

    Journal of Neuroimaging

    (1999)
  • D.J. Anderson et al.

    Batch-fabricated thin-film electrodes for stimulation of the central auditory system

    IEEE Transactions on Biomedical Engineering

    (1989)
  • E.L. Bavin et al.

    Testing auditory processing skills and their associations with language in 4–5-year-olds

    Language and Speech

    (2010)
  • X. Beebe et al.

    Charge injection limits of activated iridium oxide electrodes with 0.2 ms pulses in bicarbonate buffered saline (neurological stimulation application)

    Biomedical Engineering, IEEE Transactions

    (1988)
  • J. Bierer et al.

    Auditory cortical images of cochlear-implant stimuli: dependence on electrode configuration

    Journal of Neurophysiology

    (2002)
  • A. Boothroyd

    Profound Deafness

  • A.M. Brown et al.

    Telephone use by a multi-channel cochlear implant patient: an evaluation using open-set CID sentences

    Journal of Laryngology and Otology

    (1985)
  • G. Clark

    Cochlear Implants: Fundamentals and Implications

    (2003)
  • G.M. Clark

    The multiple-channel cochlear implant: the interface between sound and the central nervous system for hearing, speech, and language in deaf people–a personal perspective

    Philosophical Transactions of the Royal Society B

    (2006)
  • G.M. Clark et al.

    Design and fabrication of the banded electrode array

    Annals New York Academy of Sciences

    (1983)
  • G.M. Clark et al.

    A multiple-channel cochlear implant

    Acta Otolaryngology

    (1981)
  • B.M. Clopton et al.

    Estimates of essential neural elements for stimulation through a cochlear prosthesis

    The Annals of Otology, Rhinology and Laryngology

    (1980)
  • G. Cumming et al.

    Error bars in experimental biology

    The Journal of Cell Biology

    (2007)
  • P.W. Dawson et al.

    Cochlear implants in children, adolescents, and prelinguistically deafened adults

    Journal of Speech and Hearing Research

    (1992)
  • D. Deutsch

    The tritone paradox: an influence of language on music perception

    Music Perception

    (1991)
  • Cited by (10)

    • Thin-film micro-electrode stimulation of the cochlea in rats exposed to aminoglycoside induced hearing loss

      2016, Hearing Research
      Citation Excerpt :

      Thus, we conclude that when stimulating using weak currents and the TF array, the current field may have been too narrow to sufficiently activate synchronous afferent neural activity of remaining SGN. The distance of electrodes from peripheral targets was not quantified in this research and most likely had an effect on overall threshold (Allitt et al., 2012; Saunders et al., 2002). Both psychophysical and electrophysiological studies have consistently shown the distance between stimulating electrodes and the modiolus affects threshold (e.g. Allitt et al., 2012; Cohen et al., 2001).

    View all citing articles on Scopus
    View full text